|Home | About | Journals | Submit | Contact Us | Français|
The discovery that adult somatic cells can be induced to become pluripotent by overexpression of a few key transcription factors provides an exciting new window into the basic biology of pluripotency and differentiation.
The discovery three years ago that adult somatic cells can be induced to become pluripotent stem cells, so-called iPS cells, by overexpression of a few key transcription factors (Takahashi and Yamanaka, 2006), has generated much excitement because of the potential therapeutic applications of these cells (reviewed in Yamanaka, 2009). Efforts are underway in many laboratories to apply iPS cell technology to modeling diseases in vitro and to developing new tissue-replacement therapies. To fulfill the promise of iPS cells, a major focus has been to improve the efficiency and completeness of reprogramming back to a pluripotent state. Other areas of intense research include avoiding the use of viruses to deliver genes encoding the reprogramming transcription factors and hence the need for genomic integration, and deriving panels of iPS cells from patients with different diseases and targeting their differentiation in vitro into the relevant cell type. But the generation of iPS cells can also be used to gain basic insights into the biology of pluripotency and differentiation.
The mechanisms by which somatic cells are reprogrammed back to a pluripotent state are largely unknown. During the reprogramming process cells often get “trapped” in partially reprogrammed states, due in part to inefficient DNA demethylation and incomplete repression or ectopic expression of lineage-specific transcription factors (Mikkelsen et al., 2008). The contribution of each reprogramming factor is not well understood, but cMyc is thought to act early to repress somatic cell genes (Sridharan et al., 2009). Binding of the other three reprogramming factors---Oct4, Sox2 and Klf4---to pluripotency genes may be a later, rate-limiting step in the progression to complete reprogramming (Sridharan et al., 2009). These and other studies are beginning to reveal the mechanisms that underlie induction of pluripotency (reviewed in Hochedlinger and Plath, 2009). But can iPS cells provide insights into basic biology that go beyond understanding the iPS cell phenomenon?
Cells developing in vivo progress from undifferentiated states with broad cell fate potential to committed states with restricted potential. Arguably, the generation of iPS cells represents an artificial experimental manipulation that “plays the development tape backwards” and therefore may not have a parallel in vivo. However, the generation of iPS cells may involve molecular processes that have parallels with fundamental events during mammalian development (Figure 1). One such event is the reprogramming of the gamete pronuclei at fertilization, which leads to initiation of the embryonic program. The DNA in the sperm pronucleus is highly compacted and undergoes decondensation and demethylation under the influence of the oocyte's intracellular factors. Some of these same factors, most of which are unknown, are presumably also involved in the reprogramming of adult nuclei by somatic cell nuclear transfer (SCNT). The mechanisms underlying SCNT have proven difficult to dissect, mostly because of the complexity and low reproducibility of the assay.
Later in development, primordial germ cells (PGCs) also undergo a process of reprogramming that involves genome-wide demethylation of DNA and modification of histones. The generation of iPS cells from adult somatic cells involves extensive epigenetic reprogramming that includes chromatin decondensation and DNA demethylation. Epigenetic reprogramming during the generation of iPS cells may well be mediated by mechanisms very different from those that operate during in vivo reprogramming in either oocytes or PGCs, and this will need to be carefully assessed. However, should there be some molecular parallels between epigenetic reprogramming in vivo and the generation of iPS cells in vitro, then the latter may provide a particularly tractable genetic and biochemical system to dissect the underlying mechanisms. Unlike oocytes or PGCs, which exist in very limited numbers in vivo, large numbers of cells can be reprogrammed in vitro, in a quantitative and reproducible manner, to become iPS cells.
The derivation of iPS cells also may be used to explore the molecular underpinnings of germ cell tumor development. The transcriptional profile of PGCs is similar to that of embryonic stem (ES) cells (Grskovic et al., 2007), and includes expression of Oct4, Sox2, Nanog and other pluripotency-associated factors. PGCs do not express cMyc but do express high levels of another Myc family member, nMyc, and nMyc can substitute for cMyc in the generation of iPS cells (Blelloch et al., 2007). PGCs do not express Klf4, which is activated during conversion of PGCs to pluripotent stem cells in vitro. It will be interesting to determine whether acquisition of Klf4 expression, or other molecular events that occur during generation of iPS cells, play any role in transformation of PGCs and germ cell tumor development.
The derivation of iPS cells may be viewed as a biochemical assay for pluripotency. It is reproducible, quantitative and can be modulated by various conditions including the addition of cofactors (Figure 1). In this sense, genetic manipulations in the context of the iPS cell assay, such as knockdown or overexpression of candidate genes in addition to or substituting for the four reprogramming factors, may allow questions to be addressed about the molecular mechanisms that underlie pluripotency during normal embryonic development.
The initial report of iPS cell generation (Takahashi and Yamanaka, 2006) provided an important basic insight into the transcriptional networks of pluripotency. Of the four factors used by Takahashi and Yamanaka, Klf4 was the most unexpected because it is not required for pluripotency. Prompted by that initial report, a recent study shows that a redundant network of Klf2, Klf4 and Klf5 is essential for ES cell pluripotency and regulates many of the same target genes as Nanog (Jiang et al., 2008). Also, an orphan nuclear receptor required for ES cell self-renewal (estrogen related receptor-β) can substitute for cMyc and Klf4 during the induction of pluripotency (Feng et al., 2009). Thus, an interplay between research on iPS cells and ES cells may accelerate our understanding of the basic biology of pluripotency.
Another area in which the iPS cell assay may provide fundamental insights is in microRNA (miRNAs) biology. The miRNAs miR-291-3p, miR-294 and miR-295, which are expressed specifically in ES cells and regulate ES cell cycle progression, enhance the efficiency of induction of pluripotency (Judson et al., 2009). Although it remains unclear whether these miRNAs have overlapping targets in ES cells and during reprogramming, the iPS cell assay provides a complementary approach to study the biology of miRNAs associated with pluripotency.
Recent data from our laboratory suggest that some of the mechanisms that maintain the open chromatin state of ES cells may also operate during generation of iPS cells. Chd1, a chromatin remodeling factor required for maintenance of open chromatin and pluripotency of ES cells, is also required for induction of pluripotency (Gaspar-Maia et al., 2009). Therefore, the iPS cell assay can provide opportunities to gain insights into the function of transcription factors, miRNAs and chromatin remodelers in pluripotent cells. Analogous questions may be asked regarding the role of signaling proteins, metabolic pathways, cell cycle regulators, cytoskeletal proteins, etc in pluripotency.
The iPS cell assay may be particularly useful as a gain-of-function approach in the cases where loss-of-function of particular genes in early embryos or ES cells may provide only limited insight. For example, the cell cycle and metabolic pathways of early embryos and ES cells have unique properties (e.g., Wang et al., 2009). Although loss-of-function approaches to analyze these properties may lead to lethality, the iPS cell assay provides an opportunity to investigate how such properties are reassembled in somatic cells during reprogramming to the pluripotent stem cell state.
Beyond addressing these questions, the iPS cell assay may in each case provide an opportunity for structure-function studies to identify the relevant protein domains or amino acids of a particular transcription factor, chromatin remodeling protein, signaling protein, etc, that may be involved in pluripotency. It is not that such studies are impossible in early embryos or in ES cells, but rather that the iPS cell assay provides a simple complementary approach with which to accelerate the mechanistic dissection of pluripotency.
Finally, it may be possible to use variations of the iPS cell assay to gain insight into lineage commitment decisions. In a restricted number of cases, it is possible to convert one cell type into another by manipulating the expression of lineage-specific transcription factors (reviewed in Zhou and Melton, 2008). Knowledge gained about how cells change states during generation of iPS cells---for example, how they alter chromatin accessibility or reprogram epigenetic marks---may contribute to a better understanding of lineage commitment and cell-fate switching.
In addition to providing new insights into the regulation of the pluripotent state, the iPS cell assay may be a new tool to probe the molecular mechanisms that underlie the stability of the differentiated state (Figure 1). Developmental biologists have long pondered whether a terminally differentiated state is reversible and what locks that state in place. Indeed, this was the motivation behind the original SCNT experiments in amphibians (Gurdon, 1962). Several decades later, we still know very little about this research field, which may have important consequences for our understanding of cellular transformation and tumor development. Transformed cancer cells may in some cases arise by dedifferentiation, that is, by losing markers of terminally differentiated cells and acquiring stem cell-like properties. It may be possible to use the iPS cell assay to identify genes or pathways that act in differentiated cells to “lock-in” their differentiated state enabling them to resist dedifferentiation.
A recent report reveals how the iPS cell assay may be used to study the stability of the differentiated state. This report shows that inhibition of the tumor suppressor protein p53 facilitates the generation of iPS cells (Zhao et al., 2008). Other tumor suppressors, such as Rb or Pten, are also candidate repressors of dedifferentiation that may be investigated using the iPS cell assay. Lineage-specific transcription factors and epigenetic modifications such as DNA methylation or lack of H3K4 methylation at certain loci may also act to repress dedifferentiation (Mikkelsen et al., 2008). In addition, it will be of interest to test if inhibition of miRNAs expressed in differentiated cells but not in stem cells, such as let-7, can facilitate reprogramming.
More broadly, it may be possible to use the iPS cell assay to discover new genes that act as repressors of dedifferentiation in an unbiased way, for example, using large-scale RNAi screens. Screening for repressors of dedifferentiation using induction of pluripotency in different adult somatic cells may reveal some genes that act universally to repress dedifferentiation, and others that regulate the stability of specific cell types. The transcriptional similarities between pluripotent stem cells and cancer cells (Wong et al., 2008) suggest that the iPS cell assay may provide a new, quantitative approach for investigating the regulation of the differentiated cell state and how it may be subverted in cancer.
Some of these research avenues make the assumption that the same mechanisms that implement pluripotency in vivo or in ES cells operate during the generation of iPS cells in vitro. This may not necessarily be so, as exemplified by the parallel case of Mbd3, a subunit of the NuRD transcriptional repressor complex. Mouse blastocysts lacking the Mbd3 gene cannot give rise to ES cells but, when both alleles are mutated by gene targeting in preexisting ES cells, Mbd3 is not required for ES cell maintenance and propagation (Kaji et al., 2006). That is to say, Mbd3 is required for reprogramming of the inner cell mass of the embryo to the ES cell state, but not for maintenance of ES cells. It will be interesting to see if analogous examples of differences between reprogramming and pluripotency can be identified using the iPS cell assay. Presumably, there will be mechanistic similarities between the two phenomena. The effects of estrogen related receptor-β, the miR-290 family, Chd1 or blocking p53 on the generation of iPS cells, as well as the effects of the original cocktail of the four reprogramming factors, are all based on prior knowledge about the function of these various factors in ES cells.
A further caveat is that often reprogramming during iPS cell generation is incomplete and may not faithfully reproduce in vivo reprogramming or ES cell pluripotency. The appearance of incompletely reprogrammed colonies, coupled with the low efficiency of iPS cell generation, means that there is significant cell heterogeneity during reprogramming, with properly reprogrammed iPS cells constituting a small minority of the somatic cells subjected to the reprogramming factors. This is further complicated by the time it takes to observe iPS cell colonies after induction of pluripotency, which varies from 7 to 14 days in mouse cells and 14 to 28 days in human cells. These complications constitute a “black box” in which the intermediate stages of reprogramming are difficult to access. Nevertheless, current reprogramming efficiencies allow the use of end-point data, that is, the number of iPS cells properly reprogrammed can be assayed by the expression of reporter genes in a quantitative manner. Further improvements in the extent and efficiency of induction of pluripotency (Yamanaka, 2009, and references therein) will facilitate the use of iPS cells in basic biology studies. In addition, the development of readouts for early stages of reprogramming may allow the process of iPS cell generation to be visualized and studied in “realtime”.
Knowledge of the basic biology of pluripotency in early embryos and ES cells paved the way for the discovery of iPS cells. The interplay between research on iPS cells and ES cells extends beyond technical improvements to the method of iPS cell generation, and is already beginning to provide new insights into the regulation of pluripotency. The iPS cell assay may be a new quantitative tool to study the basic biology of reprogramming in vivo, and to dissect the regulation of both the pluripotent and the differentiated cell state. Research in these areas may provide new fundamental insights in developmental biology, as well as make important contributions to regenerative medicine and cancer biology.
I thank the Santos lab, R. Blelloch, J. Ramalho-Santos, M. Conti, B. Bruneau, J. Reiter and S. Fisher for helpful comments and M. Grskovic for help with the figure. The Santos lab is funded by an NIH Director's New Innovator Award, the California Institute for Regenerative Medicine and the Juvenile Diabetes Research Foundation.